Chemosphere 84 (2011) 759–766

Contents lists available at ScienceDirect

Chemosphere

journal homepage: www.elsevier .com/locate /chemosphere

Phosphate rock formation and marine phosphorus geochemistry: The deep timeperspective

Gabriel M. Filippelli ⇑Department of Earth Sciences, Indiana University – Purdue University Indianapolis (IUPUI), 723 W. Michigan St., Indianapolis, IN 46202, USA

a r t i c l e i n f o a b s t r a c t

Article history:Received 12 October 2010Received in revised form 7 February 2011Accepted 8 February 2011Available online 4 March 2011

Keywords:PhosphorusPhosphate rockOrePhosphogenesisPhosphoriteGlobal sustainability

0045-6535/$ - see front matter � 2011 Elsevier Ltd. Adoi:10.1016/j.chemosphere.2011.02.019

⇑ Tel.: +1 317 274 3795; fax: +1 317 274 7966.E-mail address: gfilippe@iupui.edu

The role that phosphorite formation, the ultimate source rock for fertilizer phosphate reserves, plays inthe marine phosphorus (P) cycle has long been debated. A shift has occurred from early models thatevoked strikingly different oceanic P cycling during times of widespread phosphorite deposition to cur-rent thinking that phosphorite deposits may be lucky survivors of a series of inter-related tectonic, geo-chemical, sedimentological, and oceanic conditions. This paradigm shift has been facilitated by anawareness of the widespread nature of phosphogenesis—the formation of authigenic P-bearing mineralsin marine sediments that contributes to phosphorite formation. This process occurs not just in continen-tal margin sediments, but in deep sea oozes as well, and helps to clarify the driving forces behind phos-phorite formation and links to marine P geochemistry.

Two processes come into play to make phosphorite deposits: chemical dynamism and physical dyna-mism. Chemical dynamism involves the diagenetic release and subsequent concentration of P-bearingminerals particularly in horizons, controlled by a number of sedimentological and biogeochemical fac-tors. Physical dynamism involves the reworking and sedimentary capping of P-rich sediments, whichcan either concentrate the relatively heavy and insoluble disseminated P-bearing minerals or providean episodic change in sedimentology to concentrate chemically mobilized P. Both processes can resultfrom along-margin current dynamics and/or sea level variations. Interestingly, net P accumulation ratesare highest (i.e., the P removal pump is most efficient) when phosphorites are not forming. Both physicaland chemical pathways involve processes not dominant in deep sea environments and in fact not oftencoincide in space and time even on continental margins, contributing to the rarity of high-quality phos-phorite deposits and the limitation of phosphate rock reserves. This limitation is becoming critical, as thehuman demand for P far outstrips the geologic replacement for P and few prospects exist for new discov-eries of phosphate rock.

� 2011 Elsevier Ltd. All rights reserved.

1. Introduction

Phosphate rocks, also referred to as phosphorites, are sedimen-tary deposits with high phosphorus (P) concentrations. These rocksare one of the primary ore sources for P (Fig. 1), which in turn is acritical and non-renewable element for fertilizer production, uponwhich global fertility depends. The startling limitation of phos-phate rock reserves has led to a renewed intensity of research inmitigating P loss from landscapes, recycling P from waste streams,and exploring new P ore potentials. From a human perspective, thissituation is one of taking an element from a high-concentrationsource and distributing it broadly—although the elemental P massis conserved, it is now scattered across the globe and effectively di-luted. The ecological impact of intense fertilization and loss from

ll rights reserved.

landscapes has been documented extensively in eutrophication lit-erature (e.g., Bennett et al., 2001; Schindler et al., 2008), but theimpacts from a resource conservation standpoint are just becom-ing apparent (Smil, 2000; Cordell et al., 2009).

From a geologic standpoint, P also is a vital fertilizing agent forbiological productivity on land and in the sea, and indeed is consid-ered the ultimate limiting nutrient in the ocean on timescalesexceeding 1000 years (Tyrell, 1999). As with all resources, geologyhad millions of years to act to form sedimentary deposits with highP concentrations, but humans have extracted these resources atrates so high that the current phosphate rock reserves might be lar-gely depleted in this century (Table 1; Cordell et al., 2009). Typicalsedimentary rocks have an average P concentration of about 0.1wt.%, whereas phosphate rocks have P concentrations 100� thatamount. Phosphate rocks are indeed unique from many perspec-tives, not least their high P content.

This paper describes the current understanding of the geochem-ical process, phosphogenesis, at the root of phosphate rock

Fig. 1. Dragline mining of phosphate ore in Florida.

Table 1Components of the Phosphorus Cycle in comparison to modern demand.

Material P content Reservoir Production/formation

(%) MT MT/yr

Average continental rocksa 0.09Marine sedimentsb 8.4 � 1011

Continental margins 0.11Phosphorites 10–20Deep Sea 0.18Hydrothermal 0.6–1.5Soilsc 0.015 40–50Oceanb �0 93 000Phosphate Rock Ored 13 2300 20Morocco and Western Sahara 820 3.4China 531 7.9USA 158 3.9Manure (cow)e 0.4 >15Sewagee 2–3 3Human demand of phosphate

rock oreHuman demand 2000e 20Human demand 2050e 30Time to reserve exhaustion 2010

demand115 years

Time to reserve exhaustion 2050demand

77 years

a Blatt and Tracy (1996).b Ruttenberg (2003).c Schlesinger (1997).d USGS (2009).e Cordell et al. (2009).

760 G.M. Filippelli / Chemosphere 84 (2011) 759–766

formation, along with the multiple geochemical and sedimentolog-ical steps that marine sediments must undergo before the rare fewultimately become phosphate rocks. A significant paradigm shift inour understanding of phosphogenesis and phosphate rock forma-tion has occurred over the past several decades, in line with in-creased understanding of the marine P cycle (Ruttenberg, 1993;Delaney, 1998; Benitez-Nelson, 2000; Colman and Holland, 2000;Compton et al., 2000; Ruttenberg, 2003). Both revolutions haveprovided a more consistent picture of phosphate rock formation,and indeed might guide the way to determine the position and ex-tent of phosphate rock reserves that are not currently economicallyviable. Various aspects of phosphorus geochemistry and phosphaterock formation have been considered before (e.g., Sheldon, 1981;Föllmi, 1996), but not from the viewpoint of critical resourcelimitation.

2. The phosphorite paradigm shift: out with the old, in with thenew

The processes responsible for phosphate rock formation havebeen the focus of research for the past century, owing to the impor-tance of this fertilizer resource. Several key early observationswere made about phosphorite deposits, especially the massivephosphorite formations that constitute the bulk of our phosphaterock reserves, called ‘‘Phosphorite Giants.’’ First, ‘‘Giants’’ containP at extremely high concentrations (Sheldon, 1981). This aloneseemed important, because in the modern ocean P is a limitingnutrient that is in such high demand by marine ecosystems thatthe element is recycled heavily and not typically found in high con-centrations in modern marine sediments. The amount of P in thePermian Phosphoria Formation, for example, was commented tobe equivalent to 5� that present in the modern ocean (Sheldon,1981). Second, the deposition of phosphorites is apparently epi-sodic through geologic time, with certain intervals (i.e., the Perm-ian, the Eocene, the Miocene) being marked by active phosphoriteformation (Cook and McElhinny, 1979; Riggs, 1984; Riggs andSheldon, 1990; Sheldon, 1989). These observations led geologiststo assume that intervals of phosphorite formation reflected uniqueintervals of Earth history where the rate of weathering and input ofP from the continents to the oceans must have been extremelyhigh to allow for ‘excess’ P burial in phosphorites. An extensionof this paradigm was that marine biological productivity and car-bon burial must have been high in these intervals of extensive Pfertilization, and thus they constituted unique climatic events aswell (Vincent and Berger, 1985).

It was with the development of better stratigraphic models, adeeper understanding of P geochemistry, and the examination ofP burial in several modern high productivity regions that the oldparadigm of atypical conditions driving deposition of ‘‘PhosphoriteGiants’’ began failing. As will be discussed in the remaining portionof the paper, the initiation of P mineralization in marine sedimentsfollows a predictable pathway seen in most (and perhaps all) mar-ine sediments. The geochemical concentration of these dissemi-nated P minerals into discrete laminae and nodules occurs inrelatively unique environments of high biological production andorganic matter burial flux with little detrital input. And finally,the concentration of these P-rich laminae and nodules into ‘‘min-able’’ phosphorites is the result of periodic disruption and winnow-ing of marine sediments caused by dynamic sedimentaryconditions which are also relatively unique. Thus, the finding ofdiscrete intervals marked by massive phosphorite deposition likelyreflects the uniqueness of these processes acting in unison, and notnecessarily the uniqueness of global conditions with respect to Pweathering and cycling. The Permian Phosphoria Formation P esti-mate was robust, but the analogy to 5� the modern marine Pinventory does not actually merit surprise given that the Phospho-ria Formation was deposited over a period of 6 million years andthus had a net P burial rate that puts it along the lines of modernupwelling environments on the Peru margin, for example (Garrisonand Kastner, 1990; Föllmi and Garrison, 1991). In so much as themodern ocean is not considered atypical, then weathering andoceanographic conditions that resulted in the deposition of thePhosphoria Formation were not atypical either. This new paradigmdoes not discount that global changes in P weathering and cyclingoccurred in the past, but simply documents that this is not a pre-requisite for phosphorite formation.

3. The marine phosphorus cycle and phosphorus sedimentation

The cycling of P in the ocean is marked by: (1) the relative pau-city of input dissolved and sediment-bound sources, (2) the role of

Fig. 2. Natural cycle of reactive phosphorus, with phosphorus input to the oceans from riverine and dust sources, output via sedimentation, and recycling via tectonics (afterRuttenberg, 2003; fluxes in Tg P/year, sedimentary reservoir in Tg P).

G.M. Filippelli / Chemosphere 84 (2011) 759–766 761

P in limiting biological productivity, and (3) the recycling and ulti-mate burial processes affecting P. Global assessments of P cyclingreveal that P weathering and release from terrestrial landscapesto the ocean is the dominant mechanism for the input of new Pto the marine environment (Fig. 2; Ruttenberg, 2003). Only a por-tion of that P is potentially bioavailable, however, and indeed muchof the P is removed in estuarine biogeochemical processes (Froe-lich, 1988). The only other input of P to the oceans is via dust,and this source likely comprises less than 10% of the total P input.Anthropogenic activities have increased the P erosion rates fromland to ocean by about 300% (Bennett et al., 2001), dwarfing the50% variation due to the extreme changes in landscapes and ero-sion during glacial cycles (Tsandev et al., 2008) and revealing thedomination of global P transfers by humans.

Phosphate concentrations are near zero in most surface waters,as this element is taken up by phytoplankton as a vital componentof their photosystems (phosphate forms the base for ATP and ADP,required for photosynthetic energy transfer) and their cells (cellwalls are comprised of phospholipids). Dissolved P has a nutrientprofile in the ocean, with a surface depletion and a deep enrich-ment. Furthermore, deep phosphate concentrations increase withthe age of deep water, and thus values in young deep waters ofthe Atlantic are typically �1.5 lM whereas those in the older Paci-fic are �2.5 lM (Broecker and Peng, 1982). Once incorporated intoplant material, P roughly follows the organic matter loop, undergo-ing active recycling in the water column and at the sediment/waterinterface.

Riverine P is present in particular and dissolved forms. Most ofthe P contained in the particulate load of rivers is held within min-eral lattices and never participates in the active biogenic cycle of P.This will also be its fate once it is delivered to the oceans, becausedissolution rates in the high pH and heavily buffered waters of thesea are exceedingly low. Roughly 50% of the total P entering theocean suffers exactly this fate (Compton et al., 2000). Of theremaining P input, some is removed via the export of organic mat-ter (with included P) to the seafloor, and a smaller portion is re-moved via scavenging by marine oxyhydroxides, especially nearocean vent environments (Feely et al., 1990; Wheat et al., 1996).Some of the particulate P is adsorbed onto soil surfaces, held with-in soil oxide, and incorporated into particulate organic matter. ThisP likely interacted with the biotic P cycle on land, and its fate upon

transfer to the ocean is poorly understood. For example, P adsorbedonto soil surfaces may be effectively displaced by the high ionicstrength of ocean water, providing an additional source of P intothe ocean. Furthermore, a small amount of the P incorporated intoterrestrial organic matter may be released in certain environmentsduring bacterial oxidation after sediment burial. Finally, some sed-imentary environments along continental margins are suboxic oreven anoxic, conditions favorable for oxide dissolution and releaseof the incorporated P. Several important studies have examined thetransfer of P between terrestrial and marine environments (e.g.,Froelich, 1988; Ruttenberg and Goñi, 1997), but more work clearlyneeds to be done to quantify the interactions between dissolvedand particulate P forms and the aquatic/marine interface. The netpre-human flux of dissolved P to the oceans is �1 Tg P/year, withan additional 1–2 Tg P/year of potentially soluble P, bringing thetotal to about 2–3 Tg P/year (Fig. 2).

Variations in the marine phosphorus cycle through time havecertainly occurred, but these variations are difficult to determinedue to a lack of effective marine proxies for the P cycle. No usefulP isotopes exist to constrain the P cycle in deep time, nor are fossil-based proxies adequate to determine past changes in the P cycle.This is due to elemental mass balance differences between P andother nutrient-type proxies—these proxies are, however, excellentfor determining the horizontal and vertical distribution of dis-solved P in ancient ocean waters. Our only direct measure of P in-put changes to the ocean, for example, is P accumulation in marinesediments. Because accumulation rates vary tremendously in dif-ferent settings, and because it is difficult to isolate P that is detritalin origin from that which has been involved in past biological pro-cesses, many important questions remain unanswered about P cy-cle variations. Nevertheless, several key events in Earth’s evolutionhave likely impacted the rate of P input to the ocean and the cy-cling of P within the ocean. For example, the Snowball Earth glaci-ations approximately 700 million years ago may have resulted in asubstantial and sustained increase in P input to the ocean, ulti-mately driving high rates of organic matter burial and the observedincrease in atmospheric oxygen critical to the evolution of animallife (Planavsky et al., 2010). Additionally, the development of ‘root-edness’ in land plants during the Devonian Era must have affectedthe rate and style of erosion of P from landscapes and input to theocean, which in turn affected marine biogeochemistry (Algeo et al.,

762 G.M. Filippelli / Chemosphere 84 (2011) 759–766

1995). Finally, a several million year long global marine biogenicbloom centering around 6 million years ago (Farrell et al., 1995),was likely the result of a P erosional pulse driven by the onset ofthe Asian Monsoon and intensification of chemical weathering ofvast floodplain and front range deposits (France-Lanord and Derry,1994) eroded from the uplifting Himalayan–Tibetan Plateau.Föllmi (1995) document many of these changes in the P cycle overthe past 160 million years, and others have examined P cyclechanges on glacial timescales (Tsandev et al., 2008; Tamburiniand Föllmi, 2009).

4. Phosphate rock formation: from phosphogenesis tophosphorite formation

The pathway from a dissolved phosphate molecule in the oceanto a P-rich phosphorite is long, and indeed is rarely travelled. Butfollowing this pathway is critical if we want to understand whatcontrols the presence and distribution of phosphate rock reservesand consider what additional P-bearing resources might be avail-able and where they might be found to fuel our P-hungry planet.

4.1. Phosphogenesis

Phosphogenesis is defined as the authigenic formation of phos-phate minerals—in marine environments, this is largely of the formcarbonate fluorapatite, or CFA. This geochemical process was longconsidered unique (the ‘‘old’’ paradigm), because our understand-ing of phosphogenesis was limited to the product (phosphate min-erals) instead of the process. Owing largely to the development byRuttenberg (1992) of a geochemical technique to isolate the vari-ous fractions of P-bearing materials in marine sediments, research-ers have now documented systematic patterns in P geochemistryand P geochemical transformations that reveal the basic processof phosphogenesis (Ruttenberg and Berner, 1993; Delaney, 1998;Schenau et al., 2000). Basically, phosphogenesis involves diagene-sis of P-bearing phases in marine sediments, the release of P tointerstitial waters, the local supersaturation of P and the authigenicformation of CFA (Fig. 3). This process can involve intermediaries,like iron oxyhydroxides, and is not 100% efficient, as P loss is seenfrom the sediments to the overlying water column (McManus

Fig. 3. A conceptual diagram of P geochemistry in oceanic sediments (after Delaney,1998). Phosphorus associated with organic matter is the primary source of P to thesediments, although the depositional flux of iron-bound P is commonly similar tothat of organic P in continental margin settings. Organic matter degradation leads tothe release of P to interstitial waters. Dissolved P in interstitial waters appears to beinvolved in several processes, including adsorption to grain surfaces, diffusion backto bottom waters, binding to iron oxyhydroxide minerals (which also scavengesome P from deep waters during particle formation in the water column), andincorporation in authigenic P, most likely as the mineral carbonate fluorapatite(CFA).

et al., 1997). But ample evidence exists to support an interpretationof authigenic CFA formation in marine sediments, spanning manyoceanographic settings.

Phosphatic shales might be considered the ‘‘type section’’ forphosphogenesis, because the CFA end-product of the process is vis-ibly apparent in these rocks (Fig. 4). In this case, phosphatic lami-nae and micronodules form authigenically from the release of Pfrom organic matter during remineralization and the migrationof this dissolved P to sites of CFA mineralization. Based on the mor-phological characteristics of this visible CFA mineralization, thisprocess seems to be controlled by a number of factors, includingthe rate of mineralization, bio-irrigation of the sediments, microbi-ological density, permeability and porosity of the sediments, thepresence or absence of viable mineral templates for CFA minerali-zation (e.g., calcite can be easily converted to CFA via phosphatesubstitution), and other competitive geochemical reactions (e.g.,Föllmi, 1996).

With the application of P geochemical techniques, however, it isclear that although phosphatic shales may be the type section forfield geologists, they are in fact just one of many environmentswhere CFA formation can be inferred. These environments rangefrom deep marine sediments, where the formation of CFA at theexpense of other P-bearing phases (so-called ‘‘sink switching’’) oc-curs on timescales of millions of years (Fig. 5), to anoxic marginalbasins where the process occurs on timescales of thousands ofyears (Fig. 5). Indeed, over the past two decades researchers haverevealed that the process of phosphogenesis is common and occursin many sedimentary environments, although it should be notedthat scant effort has been made to observe the morphology andmineralogy of the disseminated CFA that is the presumed productof this process.

Fig. 4. Phosphatic shales of the Miocene Monterey Formation (Shell Beach,California, USA). Sample A displays white nearly pure cellophane in the form ofphosphate nodules and weak lamination in an organic shale host. Sample B displaystan phosphate-rich laminations in a shale host which contains about 15% organiccarbon by mass.

Fig. 5. The progressive diagenesis of sedimentary phosphorus as a function of time in two ocean sediment cores, revealing the progressive decreases in organic-associated Pand the increase in CFA-associated P with increased time (and depth). Left—Ocean Drilling Program Site 846, eastern equatorial Pacific Ocean; Right—Ocean Drilling ProgramSite 1033 in the Saanich Inlet, British Columbia. Note that significant diagenetic transformations of P occur in less than 8000 years in the Saanich Inlet setting, driven by highorganic matter input, whereas they occur on timescales of a million years in the relatively organic poor deep sea sediments of the eastern equatorial Pacific.

G.M. Filippelli / Chemosphere 84 (2011) 759–766 763

4.2. Phosphorite formation

Marine sediments with disseminated CFA, or even those withmore concentrated occurrence of CFA as in the phosphate shaleexample, do not constitute phosphate rock reserves, at least givenmodern mining and purification economics. Indeed, the phosphaterock reserves are exclusively phosphorites, defined as a rock with Pconcentrations exceeding 9% (Blatt and Tracy, 1996). Most marinesediments have relatively constant concentrations of reactive (i.e.,non-detrital) P, whereas the P concentration in phosphorites isabout 100� higher (Table 1). Thus, the sedimentological processthat ‘‘mines’’ and concentrates widely disseminated CFA or evenmore concentrated accumulations, as in the phosphatic shaleexample above, is a critical component of phosphorite formation.

The typical marine setting for phosphorite formation is a conti-nental shelf-slope environment with extremely high surface pro-ductivity, limited dilution by terrigenous sedimentation, andperiodic winnowing and reworking. The modern analogy for thisideal environment is the Peru Margin. This setting has high coastalupwelling which produces tremendous phytoplankton blooms andhigh burial rates of organic matter with incorporated P. Terrige-nous input is limited by terrestrial dynamics—the Andes Moun-tains create a rain shadow effect for easterly storm jets, and thus

the western Andes are extremely arid resulting in little river inputinto the Pacific. Additionally, a number of off-shore basins existalong the Peru margin, with proximal basins effectively trappingwhat little terrigenous input occurs in the setting (Garrison et al.,1990). Finally, variations in sea level driven by glacial cycles driveperiodic migration of longshore and along-margin eddies. This al-lows for the accumulation of organic-rich material and relatedphosphogenesis within these sediments, followed by high currentactivity and winnowing of the lighter clay and organic materialaway from (in this case, off-shore of) the sediments and the con-centration of phosphatic materials into hardgrounds and lagdeposits characteristic of phosphorites.

This dynamic, multi-stage process of sedimentation, phospho-genesis, and winnowing, repeated many times, yields the charac-teristic sedimentology of phosphorites and results in theirextremely high P concentrations (Fig. 6). This is the truly uniqueaspect of phosphorites, and the reason why phosphorites cannotbe formed in deep sea environments, which have slow CFA forma-tion rates and typically lack the energy for concentration, or inproximal environments like the Saanich Inlet in British Columbiawhich experience too much dilution from terrigenous runoff. Somepast environments were well-suited for phosphorite formation, of-ten for different reasons, as can be seen in contrasting the major

Fig. 6. The sediment reworking process of converting sediments that have undergone high rates of phosphogenesis, as in the phosphatic shale model and example in theupper portion, into phosphorite deposits, as in the phosphorite model and the photos of off-shore phosphatic crusts (black and white photo, Blake Plateau) and the high-gradephosphorite hand sample (Bone Valley Member, Hawthorne Formation, Florida).

764 G.M. Filippelli / Chemosphere 84 (2011) 759–766

phosphatic deposits of the USA. The Miocene Hawthorne Forma-tion was deposited along the open continental shelf of the south-eastern US and experienced early diagenetic phosphogenesis, butwas the product of significant mid-Miocene sea level variationand related extensive reworking (Compton et al., 1993). The Perm-ian Phosphoria Formation of Utah, Wyoming and Idaho was depos-ited along the margins of an intracontinental seaway, whereproductivity may have been influenced by wind-blown sedimentinput and the flux of evaporative brines that pulled nutrient-richwater further into a shallow embayment setting (Hiatt and Budd,2001). The depositional setting of the Phosphoria Formation re-sulted in a number of discrete beds, related to deeper water set-tings with high concentrations of vanadium, cadmium, selenium,and uranium—age equivalent phosphorites deposited in near-shore settings are relatively free of these toxic metals (Hiatt andBudd, 2003). Finally, the phosphatic sediments of the MioceneMonterey Formation were deposited under highly anoxic condi-tions in transform sedimentary basins along the California marginthat were caused by tectonic movement related to early San An-dreas faulting. Although proximal basins trapped most of the ter-rigenous sediments, allowing for extremely pure organic mattersediments in the basins farther off-shore, the lack of reworking en-ergy in these deep basins limited the formation of true phospho-rites to a few horizons and hardgrounds—the rest of the CFAremains in laminae and nodules interbedded with organic-richshale and dolomite.

An interesting aspect of phosphorite formation is that it occursduring an interval of non-deposition or even sediment removal, viaworking or winnowing. Thus, although net P accumulation mightbe high during sedimentation and during the actual process ofphosphogenesis, it is low or even zero during times of phosphoriteformation. Thus, intervals of phosphorite formation are not neces-sarily driven by high P input from weathering or even intervals ofhigh biological production.

5. Phosphate rock reserves: a reflection on economic andenvironmental viability

Most of the reserves with the highest economic viability havebeen exploited. The clean high-quality guano phosphates fromlow-lying atolls in the central Pacific like Banaba Island and Nauruhave been completely exploited for over a decade now, although

there is some discussion of re-mining the spoil piles given the highphosphate prices today. Many of the more P-rich horizons fromstandard phosphate rocks have also been depleted, resulting inaverage P content of mined phosphate ore to decline from 15% Pin 1970 to less than 13% P by 1996 (Smil, 2002). The declining qual-ity of current reserves and the inherently low economic viability ofand environmental consequences of extraction for other potentialphosphate rock resources signals a need for a careful examinationof P usage.

In the state of Florida (USA), the well-known Bone Valley Mem-ber of the Hawthorne Formation has been mined since the late1800s. Since that time, extraction operations have not changed sig-nificantly. The main deposit lies under about 10–20 m of soil, soafter removal of this top layer, the phosphate rocks are extractedusing large draglines, and processed to purify the P and removevarious impurities (like uranium and cadmium) before final pro-cessing for the raw P fertilizer ingredient. The waste produced fromprocessing is in the form of phosphogyspsum, which contains en-ough radioactive material that it is unusable for any other purposeand is stored in large stacks and settling ponds on site. These set-tling ponds, and indeed even the piles, have inherent environmen-tal risks, as evidenced by several dam bursts which havetransported the phosphogypsum material and clays into water-ways, causing large fish kills. But in terms of resource limitation,the bottom line is that the Bone Valley Member is simply runningout. About 730 km2 of the deposit within the Peace River wa-tershed has already been mined and now is reclaimed. Miningcompanies are seeking permits to mine additional portions of themember to the south, but as this area of Florida becomes increas-ingly developed, the potential for expanded mining diminishes.This situation is the particular case for phosphate given that it isonly economic as a surface-based extraction operation. Ironically,the people who depend on a P-limited food production systemare further limiting that production by developing on top of theresource.

Along the west coast of North America the extensive MioceneMonterey Formation stretches from the Central Coast of Californiainto Baja California. The Monterey Formation hosts organic- and P-rich sedimentary rocks, and is found both on-shore and off-shore(Föllmi and Garrison, 1991). Although the phosphatic shales ofthe Monterey Formation (Fig. 2) contain a reasonable amount ofP (�6%), intervals of phosphatic hardgrounds exists within the for-mation which are classic examples of phosphorite formation due to

G.M. Filippelli / Chemosphere 84 (2011) 759–766 765

sedimentation variations and can contain up to 20% P (Föllmi andGarrison, 1991). By all measures, the P content of the MontereyFormation makes it at least a candidate for extraction should theprice of phosphate be favorable. But it is difficult to conceive ofsuch an economically favorable situation for several reasons. First,much of the on-land coastal Monterey Formation is developed.Second, off-shore-mining of phosphate rock has never been at-tempted, and thus the vast portions of the off-shore Monterey For-mation are out of reach—as are the exposed phosphatic ledges ofthe Blake Plateau and the Peru margin, for example. Finally, the ac-tive tectonics of coastal California results in the sedimentarydeposits being folded and faulted, and not flat-lying and easilyaccessible as is the case for the Bone Valley Formation of the Haw-thorne Formation.

Clearly, economics engenders innovation, and we will see tech-nologies and environmental controls in the next century that willopen up some currently uneconomic phosphate rock resources toextraction. Off-shore mining of phosphorites will be a top targetas land-based reserves are depleted and price increases. For exam-ple, off-shore Namibia and South Africa have large estimated P re-serves with very high P contents related to reworked Neogenephosphorites on the middle and outer shelf and are being targetedby for potential mining by Namibian Marine Phosphate, a jointventure between several Australian and Namibian companies(Namibian Sun, November 22, 2010). But we should be equallyclear in realizing that there is no technological ‘‘out’’ for the P prob-lem. There simply are limited resources on the earth’s surface,there is no biological replacement for elemental P, and the globalpopulation’s demand for P continues to increase with global popu-lation (Table 1; Gilbert, 2009).

6. Conclusions

Determining the mechanisms for phosphate rock formation inthe context of marine geochemistry is important to understandingthe limited availability of P-rich rock deposits on Earth. The bio-geochemical process of phosphogenesis has been found to be wide-spread in marine sediments, overturning previous concepts of itsuniqueness. But the combination of phosphogenesis plus sedimen-tologically favorably conditions for sedimentary phosphate accu-mulation plus dynamic oceanographic conditions for winnowingand concentration of P into phosphorites leads to the geographi-cally and temporally restricted occurrence of phosphate rock re-serves on Earth. Those reserves are being rapidly depleted atcurrent usage, and this likely will be accelerated with continuedglobal population growth. Given the resource limitation, drasticsteps should be taken to increase the efficiency of P use in agricul-ture, reduce the loss of P from agricultural soils, consider crops thatmore effectively mine the P that has been over-applied (e.g., peren-nial crops), and effectively recycle P from our waste streams (Smil,2002; Cordell et al., 2009). It is surprising that so much globalattention has been paid to ‘‘peak oil’’ that we have missed noticingthe threat of ‘‘peak phosphorus’’—it is time to turn that trendaround, as we will have to wait tens of millions of years for geolog-ical processes to restore the concentrated P that we have dissemi-nated in two centuries.

Acknowledgements

I would like to thank a fantastic set of colleagues and organizerswho participated in the Phosphorus Cycle workshop of the AspenGlobal Change Institute for introducing me to the P scarcity litera-ture and for being interested in the geologic aspects of phosphatereserve formation. Reviews by John Compton, Eric Hiatt, and David

Vaccari were extremely helpful. I acknowledge research supportfrom the NSF (OCE9711957, BCS9911526, OCE0452428).

References

Algeo, T.J., Berner, R.A., Maynard, J.P., Scheckler, S.E., 1995. Late Devonian oceanicanoxic events and biotic crises: ‘‘Rooted’’ in the evolution of vascular landplants? GSA Today 5, 64–66.

Benitez-Nelson, C.R., 2000. The biogeochemical cycling of phosphorus in marinesystems. Earth Sci. Rev. 51, 109–135.

Bennett, E.M., Carpenter, S.R., Caraco, N.F., 2001. Human impact on erodiblephosphorus and eutrophication: a global perspective. Bioscience 51, 227–234.

Blatt, H., Tracy, R.J., 1996. Petrology, second ed. Freeman, New York.Broecker, W.S., Peng, T.-H., 1982. Tracers in the Sea. Eldigio Press, Palisades, New

York.Cook, P.J., McElhinny, M.W., 1979. A reevaluation of the spatial and temporal

distribution of sedimentary phosphate deposits in the light of plate tectonics.Econ. Geol. 74, 315–330.

Colman, A.S., Holland, H.D., 2000. The global diagenetic flux of phosphorus frommarine sediments to the oceans: Redox sensitivity and the control ofatmospheric oxygen levels. In: Glenn, C., Prévôt-Lucas, L., Lucas, J. (Eds.),Marine Authigenesis: From Microbial to Global, vol. 66. SEPM SpecialPublication, pp. 53–76.

Compton, J., Mallinson, D., Glenn, C., Filippelli, G., Föllmi, K., Shields, G., Zanin, Y.,2000. Variations in the global phosphorus cycle. In: Glenn, C., Prévôt-Lucas, L.,Lucas, J. (Eds.), Marine Authigenesis: From Microbial to Global, vol. 66. SEPMSpecial Publication, pp. 21–34.

Compton, J.S., Hodell, D.A., Garrido, J.R., Mallinson, D.J., 1993. Origin and age ofphosphorite from the south-central Florida Platform: Relation ofphosphogenesis to sea-level fluctuations and d13C excursions. Geochim.Cosmochim. Acta 57, 131–146.

Cordell, D., Drangert, J.-O., White, S., 2009. The story of phosphorus: Global foodsecurity and food for thought. Global Environ. Change 19, 292–305.

Delaney, M.L., 1998. Phosphorus accumulation in marine sediments and the oceanicphosphorus cycle. Global Biogeochem. Cycle 12, 563–572.

Farrell, J. et al., 1995. Late Neogene sedimentation patterns in the eastern equatorialPacific Ocean (Leg 138). In: Blum, P., Gersonde, R., Hodell, D. (Eds.), Proceedingsof the Ocean Drilling Program, Scientific Results, vol. 138. College Station, Texas,pp. 717–756.

Feely, R.A., Massoth, G.J., Baker, E.T., Cowen, J.P., Lamb, M.F., Krogsland, K.A., 1990.The effect of hydrothermal processes on midwater phosphorus distributions inthe northeast Pacific. Earth Planet Sci. Lett. 96, 305–318.

Föllmi, K.B., 1995. 160 m.y. Record of marine sedimentary phosphorus burial:Coupling of climate and continental weathering under greenhouse and icehouseconditions. Geology 23, 859–862.

Föllmi, K.B., 1996. The phosphorus cycle, phosphogenesis and marine phosphate-rick deposits. Earth Sci. Rev. 40, 55–124.

Föllmi, K.B., Garrison, R.E., 1991. Phosphatic sediments, ordinary or extraordinarydeposits? The example of the Miocene monterey formation (California). In:Muller, D.W. et al. (Eds.), Controversies in Modern Geology. Academic Press,London, pp. 56–84.

France-Lanord, C., Derry, L.A., 1994. D13C of organic carbon in the Bengal fan: sourceevolution and transport of C3 and C4 plant carbon to marine sediments.Geochim. Cosmochim. Acta 58, 4809–4814.

Froelich, P.N., 1988. Kinetic control of dissolved phosphate in natural rivers andestuaries: a primer on the phosphate buffer mechanism. Limnol. Oceanogr. 33,649–668.

Garrison, R.E., Kastner, M., 1990. Phosphate sediments and rocks recovered from thePeru margin during ODP Leg 112. In: Suess, E., von Huene, R. (Eds.), Proceedingsof the Ocean Drilling Program, Scientific Results, vol. 112. College Station, Texas,pp. 111–134.

Garrison, R.E., Kastner, M., Reimers, C.E., 1990. Miocene phosphogenesis inCalifornia. In: Burnett, W.C., Riggs, S.R. (Eds.), Phosphate Deposits of theWorld. Neogene to Modern Phosphorites, vol. 3. Cambridge University Press, pp.285–299.

Gilbert, N., 2009. The disappearing nutrient. Nature 461, 716–718.Hiatt, E.E., Budd, D.A., 2001. Sedimentary phosphate formation in warm shallow

waters: new insights into the paleoceanography of the Permian Phosphoria Sea.Sediment. Geol. 145, 119–133.

Hiatt, E.E., Budd, D.A., 2003. Extreme paleoceanographic conditions in a Paleozoicoceanic upwelling system: organic productivity and widespreadphosphogenesis in the Permian Phosphoria Sea. In: Chan, M.A., Archer, A.W.(Eds.), Extreme Depositional Environments: Mega End Members in GeologicTime, vol. 370. Geological Society Special Paper, pp. 245–264.

McManus, J., Berelson, W.M., Coale, K.H., Johnson, K.S., Kilgore, T.E., 1997.Phosphorus regeneration in continental margin sediments. Geochim.Cosmochim. Acta 61, 2891–2907.

Planavsky, N.J., Rouxel, O.J., Bekker, A., Lalonde, S.V., Konhauser, K.O., Reinhard, C.T.,Lyons, T.W., 2010. The evolution of the marine phosphate reservoir. Nature 467,1088–1090.

Riggs, S.R., 1984. Paleoceanographic model of Neogene phosphorite deposition, USAtlantic continental margin. Science 223, 123–131.

Riggs, S.R., Sheldon, R.P., 1990. Paleoceanographic and paleoclimatic controls of thetemporal and geographic distribution of Upper Cenozoic continental margin

766 G.M. Filippelli / Chemosphere 84 (2011) 759–766

sediments. In: Burnett, W.C., Riggs, S.R. (Eds.), Phosphate Deposits of the World.Neogene to Modern Phosphorites, vol. 3. Cambridge University Press, pp. 53–61.

Ruttenberg, K.C., 1992. Development of a sequential extraction method for differentforms of phosphorus in marine sediments. Limnol. Oceanogr. 37, 1460–1482.

Ruttenberg, K.C., 1993. Reassessment of the oceanic residence time of phosphorus.Chem. Geol. 107, 405–409.

Ruttenberg, K.C., 2003. The global phosphorus cycle. In: Schlesinger, W.H. (Ed.),Treatise on Geochemistry, vol. 8. Elsevier, NY, pp. 585–643.

Ruttenberg, K.C., Berner, R.A., 1993. Authigenic apatite formation and burial insediments from non-upwelling continental margin environments. Geochim.Cosmochim. Acta 57, 991–1007.

Ruttenberg, K.C., Goñi, M.A., 1997. Phosphorus distribution, C:N:P ratios, and d13Coc

in arctic, temperate, and tropical coastal sediments: Tools for characterizingbulk sedimentary organic matter. Mar. Geol. 139, 123–146.

Schenau, S.J., Slomp, C.P., De Lange, G.J., 2000. Phosphogenesis and activephosphorite formation in sediments from the Arabian Sea oxygen minimumzone. Mar. Geol. 169, 1–20.

Schindler, D.W., Hecky, R.E., Findlay, D.L., Stainton, M.P., Parker, B.R., Paterson, M.J.,Beaty, K.G., Lyng, M., Kasian, S.E.M., 2008. Eutrophication of lakes cannot becontrolled by reducing nitrogen input: results of a 37-year whole-ecosystemexperiment. Proc. Natl. Acad Sci. USA 105, 11254–11258.

Schlesinger, W.H., 1997. Biogeochemistry: An Analysis of Global Change. Acad.Press, San Diego.

Sheldon, R.P., 1981. Ancient marine phosphorites. Annu. Rev. Earth Planet Sci. 9,251–284.

Sheldon, R.P., 1989. Phosphorite deposits of the Phosphoria Formation, westernUnited States. In: Notholt, A.J.G. et al. (Eds.), Phosphate Deposits of theWorld. Phosphate Rock Resources, vol. 2. Cambridge University Press, pp. 53–61.

Smil, V., 2000. Phosphorus in the environment: Natural flows and humaninterferences. Annu. Rev. Energy Environ. 25, 53–88.

Smil, V., 2002. Phosphorus: global transfers. In: Douglas, P.I. (Ed.), Encyclopedia ofGlobal Environmental Change. John Wiley & Sons, Chichester, pp. 536–542.

Tamburini, F., Föllmi, K.B., 2009. Phosphorus burial in the ocean over glacial-interglacial time scales. Biogeosciences 6, 501–513.

Tsandev, I., Slomp, C.P., Van Cappellen, P., 2008. Glacial-interglacial variations inmarine phosphorus cycling: Implications for ocean productivity. GlobalBiogeochem. Cycle 22, GB4004.

Tyrell, T., 1999. The relative influence of nitrogen and phosphorus on oceanicprimary productivity. Nature 400, 525–531.

USGS, 2009. Mineral Commodity Summaries: Phosphate Rock, United StateGeological Survey. Reston, VA.

Vincent, E., Berger, W.H., 1985. Carbon dioxide and polar cooling in the Miocene:the monterey hypothesis. In: Sundquist, E.T., Broecker, W.S. (Eds.), The CarbonCycle and Atmospheric CO2: Natural Variations Archean to Present, vol. 32.American Geophysical Union Monograph, pp. 455–468.

Wheat, C.G., Feely, R.A., Mottl, M.J., 1996. Phosphate removal by oceanichydrothermal processes: An update of the phosphate budget of the oceans.Geochim. Cosmochim. Acta 60, 3593–3608.